Clinical outcomes and complications after anterior cruciate ligament reconstruction with bone–patellar tendon–bone in patient Tanner 3 and 4: a systematic review

Background Clinical outcomes and potential complications associated with Bone–Patellar Tendon–Bone (BPTB) graft in skeletally immature ACL reconstruction (ACLR) are poorly defined. Considering that in Tanner 1–2 patients this kind of graft is not recommended, we focused our systematic review on the evaluation of all the studies in the literature that reported clinical outcomes and rate of complications of the ACLR using BPTB graft in Tanner 3–4 patients. Methods This review was conducted in accordance with the PRISMA statement. PubMed, Cochrane Library, EMBASE and Scopus were examined from 1965 to 2020 using different combinations of the following keywords: “ACL reconstruction”, “skeletally immature”, “young”, “patellar tendon” and “BPTB”. The database search yielded 742 studies, on which we performed a primary evaluation. After carrying out a full-text evaluation for the inclusion criteria, 4 studies were included in the final review and assessed using the Newcastle–Ottawa scale. Ninety-six cases with mean age of 14.2 years were reported. Results Good stability and functional outcomes were reported with a mean follow-up of 49.5 months. Return to sport rate ranged from 91.7% to 100%. A KT-1000 side-to-side difference higher than 5 mm was observed in five patients (5.2%). No lower limb length discrepancy and angulation were reported. Graft rupture rate was 5.2%. Conclusion According to these results, BTPB graft could be a good choice in Tanner 3–4 patients who want to achieve their preinjury sport level with a low risk of growth disturbances and graft failure. Further investigations in a wider population are needed.


Introduction
Anterior cruciate ligament (ACL) injury is common in children and adolescent and its incidence is getting higher year by year [1,2]. In a population between 6 and 18 years old, the incidence is 121/100,000 per year, with slight differences between male (114/100,000) and female (129/100,000). ACL injury is also one of the most common injuries in paediatric population representing 6.7% of all injuries and 30.8% of all knee injuries in soccer players between 5 and 18 years old [3,4]. Age is an important ACL tear risk factor with an average rate of incidence that increased by 2.3% every year, reaching peak in 16-year-old females and in 17-year-old males [5]. Sports that require cutting movements like football, soccer or basketball are at higher risk of ACL injury too with an incidence that varies between 1 and 3.4% [6].
ACL deficient knees are related to poor long-term outcomes including low objective IKDC scores, increased anterior tibial translation at arthrometry and radiological evaluation, increased joint laxity and extensive arthritic changes in the injured leg [7]. McCarroll et al. found that adolescents with ACL deficiency treated conservatively experienced recurrent instability, effusion and pain during activities [8]. Moreover, an ACL deficient knee often leads to secondary meniscal and/or cartilage damage, which may lead to knee degeneration and functional instability [9].
Different approaches are available in skeletally immature patients to restore ACL function and prevent potential damage of the growth plates [10,11]. Physeal sparing surgical techniques are the most common among open physis patients including all-epiphyseal and extra-articular reconstruction [12], partial transphyseal and transphyseal reconstruction with soft tissue graft are mainly used in patients with low growth potential [13]. However, recent data suggest that drilling across open proximal tibial or distal femoral physes can be a safe and effective procedure in patients close to skeletal maturity. Animal studies show that risk of growth plate disturbance is related to tunnels of 7-9% of the cross-sectional area of the physis. Small and centrally placed tunnels are recommended to a minimizes the risk of physeal closure [14]. Moreover, transtibial techniques reduce the risk of femoral growth plate violations compared to anteromedial approach that could produce an elliptical tunnel with a larger and more lateral growth plate violation [15].
In immature patients, the main concern about growth plates is the risk of bone bridging and deformities. Considering that, some surgeons prefer soft tissue graft as hamstring for ACLR in skeletally immature patients [16][17][18][19]. An MR imaging study after transphyseal reconstruction of the ACL in skeletally immature adolescent patients shows a focal bone bridge in 11% of patients. However, no growth disturbances were observed in these patients [20].
Despite the large amount of research about ACL injury in skeletally immature patients, the choice regarding the best kind of graft for reconstruction is still widely debated [21]. Ligamentization process, long-term results and risk of rerupture are not well defined yet. However, a high re-rupture risk after reconstruction with soft tissue graft in paediatric patients was reported in selected techniques [22][23][24]. Low re-rupture rate was reported in adult ACLR with BPTB, however clinical results, lower limb discrepancy rate and re-rupture rate of BPTB ACLR in young population are not well defined [25].
The aim of our systematic review is to evaluate all the studies in the literature that reported clinical outcomes and rate of complications of the ACLR using BPTB graft in almost skeletally mature patients.

Focused question based
Based on the Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) guidelines [26], we stated three specific questions: (1) What are the clinical outcomes of ACLR using BPTB in skeletally immature patients? (2) What is the reinjury rate after this technique? (3) How many complications are reported after ACLR using BPTB graft in skeletally immature patients?

Eligibility criteria
The following inclusion criteria were used to determine study eligibility: (i) original clinical studies, (ii) case-control and cohort study, (iii) skeletally immature patients, (iv) patients with ACL injury and (v) type of intervention: transphyseal ACLR using patellar tendon (Fig. 1). Letters to the editor, historic reviews, case reports, case-series and unpublished articles were excluded.
Titles and abstracts of studies identified were screened by two authors independently. If they met the eligibility criteria, they were evaluated for their full text. A search using ResearchGate was done in order to evaluate papers reported in sections "references" and "citations" of potentially relevant original articles found during the previous step. Finally, a search using "similar articles" was done for papers selected. Studies found were discussed by the authors before including them into the review.

Study selection
The search results are shown in Fig. 2, according to PRISMA guidelines. The search identified 742 results of which 738 did not satisfy the eligibility criteria and so they were excluded. Studies in which the reconstruction with patellar tendon was reinforced with extra-articular procedure, such as the one by McCarrol et al. [8], were excluded from the review too. In total four studies were considered and processed for data extraction.

Methodological study quality assessment
The Newcastle-Ottawa scale [27] (NOS) was used to grade the methodological quality of each study assessed in this review (Table 1). In summary, the NOS scale uses a systematic approach based on 3 specific criteria: Selection (S), Comparability (C) and Exposure (E), which are subdivided in 9 criteria: (S1) representativeness of the exposed cohort; (S2) selection of the non-exposed cohort; (S3) ascertainment of exposure; (S4) demonstration that outcome of interest was not present at start of study; (C1) comparability of cohorts; (C2) controls on the basis of the analysis; (E1) assessment of outcome; (E2) follow-up long enough for outcomes to occur; (E3) Adequacy of follow-up of cohorts. Each study could have a maximum score of 9.

Study characteristics
Main features of every study are summarized in Table 2. All the studies were cohort studies, only one of them was prospective [28], the others were retrospective [1,28,29]. The number of patients included in the studies ranged from ten to 60, with a total amount of 96 cases, of which 54 were males, while 42 were females. The mean age at surgery

Main outcomes
The outcomes are summarized in Table 3 [29,30] used both clinical and radiological tests. Neither growth plate disturbances and arrests [29,30], nor evidence of varus or valgus angulation and limb length discrepancy [1,[28][29][30] were reported. No one found any alteration of growth, the height increasing after surgery was reported by three studies [1,29,30]

Secondary outcomes
All the studies reported a high rate of return to sport ranging from 91.7 to 100% [1,[28][29][30]. Patients returned to their preinjury level of daily activity and athletic participation.
Different patient-reported outcome measures (PROMs) were used to evaluate knee conditions. Two studies adopted the IKDC subjective knee evaluation form [1,30] good and one fair results with an average score of 87.7 [28]. No paediatric PROMs were utilized.
KT-1000 was performed by every study [1,[28][29][30] in order to test knee instability after ACLR [31]. All the authors reported good results. Only five patients (5.2%) belonging to the studies of Memeo et al. and McCarroll et al. presented a KT-1000 side-to-side difference higher than 5 mm [28,29].

Discussion
Historically, the most important concerns about transphyseal techniques were related to physeal damages and growth disturbances [9,32]. Generally, soft tissue grafts are recommended in skeletally immature patients, however BPTB graft are occasionally utilized also in young patients [33].
The main finding of this study is that transphyseal ACLR with BPTB in patients who have almost achieved skeletal maturity is a technique with reported good clinical outcomes with low graft rupture rate without growth disturbances in a selected population.
Indeed, in our review of all the studies, no evidence of growth disturbances, neither in clinical nor in radiological test were reported. In particular, authors highlighted the absence of lower limb discrepancy, early physeal arrest, varus or valgus angulation, medial or lateral joint space narrowing.
However, an MRI study concerning focal bone bridge after this technique was not already performed. The only physical alteration is a lack of extension occurring two patients reported by McCarroll et al. in 1994 [29]. More recent studies did not describe this complication. This statement may be linked to different factors. First of all, McCarrol et al. performed surgery using prevalently an open procedure by mini arthrotomy or through the patellar defect [29]. In fact, only three patients out of 60 underwent an arthroscopically assisted procedure. Secondary, more recent studies can rely on improvements in BPTB technique and better experience of the surgeons due to the increasing rate of ACLR in the skeletally immature patient. These aspects may have led to fewer complication rate of surgical procedure.
We believe that a meticulous growth potential evaluation is another essential point to plan the adequate surgical technique: chronological age, skeletal age, knee growth plates maturity evaluation and Tanner staging should be considered in the selection of the graft and of the surgical technique [13,34,35].  Re-injury rate: 5% respectively) [30]. Memeo et al. performed an evaluation of the maturity before the surgery through the assessment of Tanner staging, X-rays and MRI. All patients presented with Tanner stage 3 with radiological evidence of open tibial and femoral physes [28]. McCarrol et al. evaluated preoperatively Tanner staging, adolescent growth spur, height and X-rays to determine if the patients could be considered skeletally mature. Another important feature which influences the choice of the graft is the failure rate. In our review the authors reported five graft failures out of 96 patients treated, it means 5.2%. This value suggests that patellar tendon graft is also less likely to fail compared to the overall rate of ACL graft failure in children and adolescent patients. According to Ho et al., in fact, graft failure was identified in 9.6% of the 561 patients who underwent ACLR [36]. Results achieved by Ho et al. reflect the ones found in our review. They reported similar incidence of patellar tendon graft failure, 6%, confirming that this type of graft has the lowest failure rate in this kind of population. Patients treated with soft tissue graft, in fact, had more than twice the probability (13%) to have a re-rupture. These findings are confirmed by different studies on hamstring graft including the ones by Cohen et al. [37], Calvo et al. [38] and Pennock et al. [39] which reported a failure rate of 11, 12 and 21, respectively.
All the authors reported good results in terms of anterior-posterior stability considering KT-1000 side-to-side difference [1,[28][29][30]. Studies included in our review reported 91 KT-1000 side-to-side difference lower than 5 mm and 5 higher [29]. Thus, 5.2% of ACLR using patellar tendon generate a certain degree of laxity. Anyway, this is not necessarily related to symptomatology, in fact, McCarroll et al. stated that no one of the three athletes complained episodes of giving way [29].
Wong et al. in their meta-analysis of ACL rupture in skeletally immature subjects reported the average result of IKDC and Lysholm score obtained from 23 and 20 studies, respectively [40]. The IKDC score ranged from 81 to 100 with 88% of grade A or B, the Lysholm score had a mean value of 94.6 points. In our review, we reported similar outcomes, however a prospective controlled group is recommended to better compare different techniques in skeletally immature patients.
Patellar tendon graft has been described by Memeo [1,28,30]. According to Kay et al. [41], the overall rate of return to preinjury level after ACLR is 78.6%, while in our review the return to sport using patellar tendon is 93.8%. Actually, the range of age evaluated in the metaanalysis is wider, considering children from 6-year-old, but the mean age is perfectly comparable with the one of the studies included in our review.
The review has some limitations. The first one concerns the sample, and it can be divided into two topics: quantity and quality. Quantity: the limited number of patients enrolled by the four studies included in the review, only 96, might not be large enough to represent the entire population. Quality: different choices of graft (allograft vs autograft) and different procedures (arthrotomy vs arthroscopy) were performed so it could dilute the methodology. However, considering that all the papers showed similar good results, independently from the graft or the procedure performed to reconstruct the ACL, we believe that these differences do not affect the safeness of BPTB technique. A second limitation is due to the lack of recent studies about the topic, in fact, papers analysed were published from 1994 to 2012, therefore the results achieved might be obsolete because of changing in different aspects such as: surgery, post-op rehabilitation, follow-up and prevention of re-injury.
Another limitation is due to the age of the patients enrolled. Since the mean age at surgery was 14.2 years it is difficult to establish the real occurrence of growth abnormalities. Having younger patient would grant better confidence for this objective. Unfortunately, the literature is lacking in studies about ACL reconstruction with BPTB technique in the youngest.
The last limitation is represented by the design of the studies included in the review, in fact there are no control groups, and all the data were collected in a retrospective way.

Conclusion
Due to increasing involvement in highly competitive sport, the amount of ACLR in children and adolescents is getting higher. Surgery goal is to restore the preinjury activity level in the safest way possible with a low rate of ACL graft re-rupture. Patellar tendon graft could be a good choice in patients Tanner 3 and 4 with this aim. In fact, good clinical outcomes with a low graft rupture rate were reported without growth disturbances in this selected population. Further clinical and instrumental studies are recommended to better understand the real role of BPTB graft in a wider cohort of skeletally immature patients. Availability of data and materials Manuscript has not associated data in a data repository. If you need information please contact the corresponding author at the following address: marco.turati@unimib.it.

Conflict of interest:
None of the authors have any conflicts of interest to declare.
Ethical approval Not required.
Consent to participate Not applicable.

Consent to publish Yes.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.